Cloud TPUs in Beta

Re:Cloud TPUs in Beta by jlebar on Monday February 12, 2018 @12:26PM (#16360156)

I know people don't know what to expect from tpu performance, but does anyone actually get 100tops out of Volta? I thought you'd have to spin the tensorcores and never touch memory, which is...not realistic.

I know you hedged by saying "back of the envelope", but I'd much rather compare on real benchmarks than based on cited peak performance numbers, which are kind of meaningless.

[ Parent | on HN ]

Re:Cloud TPUs in Beta by deepnotderp on Monday February 12, 2018 @01:14PM (#16360556)

This is true of the TPU as well, check out their paper's utilization numbers. If you ignore one outlier at ~90% utilization, their utilization plummets. I'm glad people are finally looking past the b.s. "peak" numbers for once though.

[ Parent | on HN ]

Re:Cloud TPUs in Beta by jlebar on Monday February 12, 2018 @01:48PM (#16360871)

Has Google published data on the memory bandwidth of TPU v2 (aka "cloud TPU")? I'm having trouble finding it.

In any case I agree, we shouldn't be looking at the stated peak compute of either of the chips.

(Disclaimer: I work at Google on XLA, and have in the past worked on TPUs.)

[ Parent | on HN ]

Re:Cloud TPUs in Beta by boulos on Monday February 12, 2018 @02:09PM (#16361075)

From the blog post is the link to the fairly recent NIPS presentation: https://supercomputersfordl2017.github.io/Presentations/Imag...

which claims 2400 GB/s for the board and 600 GB/s per “chip”.

[ Parent | on HN ]

Re:Cloud TPUs in Beta by jlebar on Monday February 12, 2018 @02:31PM (#16361297)

This is in comparison to 900gb/s for V100.

[ Parent | on HN ]